It is well known that melt grown ingots such as are used in the production of commercially acceptable optical bodies, must pass the most stringent quality control requirements. In fact, the continuing demand for better performance of optical bodies requires that ingots be grown as nearly perfectly as possible. Since impurities in the melt generally degrade performance as an optical body, it is essential that growth stock for a melt-grown ingot be as pure as possible. By this is meant that deleterious impurities in crystal stock be less than 1 part per million (ppm). Even so, it often happens that stock which passes the most stringent purity specification still unpredictably yields an ingot which is unacceptable from the standpoint of color, afterglow or hardness.
By "unacceptable color" we refer to the photosensitivity of an ingot as evidenced by solarization. By "solarization" we refer to darkening of the crystal when it is exposed to light. Crystals which resist solarization are deemed to have acceptable color, that is a water-white color. In fact, an ingot may be removed from the crucible in which it is grown, and appear to have perfectly acceptable color, yet, after it is exposed to ultraviolet light for a short period of time, the ingot is visibly darkened. Such darkening of an ingot, though it may fade in time at room temperature, indicates the presence of impurities which may be present in so small a concentration as to be undetectable by any of the conventional analytical means, whether chemical or physical. An ingot with a color problem is scrapped. Since the cost of scrapping even a relatively small ingot is substantial, it is unnecessary to state that one does everything possible to avoid the cost of scrapping an ingot in excess of 20 inches in diameter, which cost can easily exceed several tens of thousands of dollars.
By "unacceptable afterglow" we refer to a scintillation phosphor ("scintillator") which upon excitation produces light pulses over a period substantially more than 10 microsecs, at a level of intensity which interferes with the measurement of light output of subsequent pulses. A low quality scintillator may exhibit an afterglow for many seconds, and even minutes.
By "unacceptable hardness" we refer to a noticeable increase in hardness of the clear portion of the ingot, due for example, to the distribution of calcium impurity, which makes a difference in the type of surface finish which may be imparted the usable portion of the ingot. The type of finish alters the reflectance of a sanded, machined or burnished (say, with steel wool), surface of scintillator units fabricated from the ingot.
This invention does not purport to permit the substitution of relatively impure charge stock for the high-purity charge stock required in the growth of scintillator ingots, but it does provide a solution to the aforedescribed problems which unexpectedly arise even when high-purity or ultra-pure stock is used. A typical charge stock for a melt-grown ingot may be commercially available crystals of ultra-purity, or stock obtained by purification of commercially available, less pure material, as described for example in U.S. Pat. No. 2,640,755 to J. Hay. Such virgin crystalline charge stock typically used for growing ingots, is referred to as "fresh powder stock". Charge stock may also constitute scrap obtained from cutting useful scintillators from an acceptable ingot. Such scrap, after inspection to reject pieces having visible inclusions of foreign material, is referred to as "remelt scrap" and is generally charged, after it is crushed, to a crucible. In some cases, large pieces of remelt scrap can be stacked in the crucible provided their shape does not allow the charge to shift during melt-down.
As is well known, the use of getters or scavengers in the growth of melt-grown ingots involves a mechanism particularly noted for its unpredictability. Typically, a getter is used to remove a specific impurity known to be present in a particular melt-grown ingot. For example, a trace of free bromine in the atmosphere above a melt of an alkali metal chloride or alkali metal bromide, is disclosed in U.S. Pat. No. 4,055,457. The bromine serves to suppress sulfate, nitrate and nitrite ions. In another example, potassium chloride ingots are grown in the presence of carbon tetrachloride in the atmosphere, which CCl.sub.4 at growth temperature provides phosgene to scavenge oxygen (see Pastor, R. C. and Braunstein, M., Air Force Weapons Laboratory Report AFWL-TR-72-152, Vol. II, p 103-108, July 1973).
As is also well-known, scintillator ingots are grown in fused silica crucibles, particularly when water is excluded by an inert gas sweep, as taught by Lafever, R. A. in U.S. Pat. No. 2,984,626. It is possible that the melt is inadvertently contaminated with silica if the temperature of the melt is high enough, and/or the charge contains an impurity that is alkaline, or generates a basic reaction. Whether or not this silica contaminant is deemed active, there is no evidence of the extent, if any, of such possibly beneficial contamination. Ingots grown in fused silica crucibles from pure growth stock, in the absence of moisture and with an inert gas sweep, are indistinguishable from those grown from the same charge stock in platinum crucibles, all other conditions of growth being the same.
We have previously used silica (SiO.sub.2) alone in a reactive form, as a useful getter for a melt from which a scintillator ingot is grown. However, silica alone, in an amount in the range from about 10 ppm to about 100 ppm, produces an undesirably high amount of "floc". This cotton-like floc is unavoidably retained as inclusions in scintillator units fabricated from the ingot. More importantly, the use of active SiO.sub.2 alone requires a relatively high "soak temperature" at which the melt is superheated during the period before crystal growth to make the SiO.sub.2 react with the deleterious trace impurities, and also to melt silicates (disilicates) formed by the reaction with active silica alone. The soak temperature is generally greater than 100.degree. C. above the melting point of the charge stock. With addition of reactive oxides of boron (hereafter "borate" for brevity), the soak temperature is generally less than about 200.degree. C., and preferably less than about 100.degree. C. above the melting point of the charge stock. Borate addition, by itself, in the absence of active silica, shows no appreciable change in performance or reduction in sensitivity.
Minor variations in a method of growing an ingot often result in a substantially different optical body, whether it be for the better or for the worse. At the very low concentrations of impurities which prove to be deleterious, contamination from the furnace becomes a surprisingly important factor when growing an ingot from ultra-high purity stock. Heretofore, improvements in ingot quality were sought by minor alterations in the conditions of growth, and/or a beneficial, if unpredictable, stratification of the ingot in such a way as to concentrate the impurities in a portion of the ingot which can be discarded. However, where concentration of deleterious impurities is so low as to be undetectable by conventional chemical or spectrographic means, it is most difficult to concentrate the impurity in any particular portion of the ingot, or to trace its origin. In other words, an effective getter must remove the effects of the deleterious impurities no matter what their origin or how they are distributed in the ingot, which requires that the getter be able to combat the effects of a wide range of contaminants.
Briefly stated, when metal or non-metal trace impurities are distributed throughout a melt, an effective getter must: (1) be dispersed throughout the melt, and essentially homogeneously distributed for Stockbarger growth where melt-stirring is minimal; (2) react with the impurities present without removing too much of a dopant or activator deliberately added to the melt; (3) tie up the reaction products in such a way as not to affect the optical performance of the finished ingot; and yet (4) have characteristics such that the presence of the getter in the melt-grown ingot is not objectionable. These many exacting requirements are satisfied to our knowledge, only by the combination of reactive oxides of boron and silicon, (thus, referred to as a "combination getter"), or a compound which yields one or the other, or both, of the desired reactive oxides. The oxides of titanium, aluminum, zirconium, lanthanum, gallium, tin, lead and other members of Groups III and IV of the Periodic Table, are ineffective getters, if not deleterious contaminants, in an alkali metal halide melt.
A melt-grown alkali metal halide ingot grown according to the process of this invention in which a combined getter of borate and active silica has been used, has most of the floc formed by reaction of the getter with impurities in the melt, distributed near either the upper surface of the ingot for some melts, or the bottom surface for other melts; and, some floc distributed around the sides of the ingot. Most of the floc is rejected by the melt, before or during growth of the ingot, in such a way that the portions of the ingot containing most of the floc can be discarded. The floc within the melt is characteristically present therein as metal silicate and metal borate solid particles or liquid droplets which tend to cluster or agglomerate. Excess unreacted silica is insoluble in alkali metal halide melts, and collects along with the metal borosilicate floc. This floc tests high in boron and various metals not found in analysis of the growth stock, but the floc may or may not contain all of the added borate component of the getter, some of the borate ends up in solid solution in the crystal when the procedure described herein is followed. Such an ingot, grown from a treated melt, fails to exhibit visible darkening upon exposure to ultraviolet light in an amount which would darken an otherwise identical ingot which was not grown from a melt which was treated (hence referred to herein as "treated melt" or "treated ingot"), with the combination getter.